- THE OAA
- NEWS & EVENTS
- PROFESSIONAL RESOURCES
- PUBLIC RESOURCES
- DISCOVER AN ARCHITECT
Recently, as part of a retrofit process to improve energy efficiency, the OAA had its main office building at 111 Moatfield Drive in Toronto tested for airtightness. A team from Building Science Corporation, working for Halsall Associates, installed an array of electronically controlled blower doors and proceeded to conduct a series of airtightness tests to determine how much air leaks through the building enclosure. This article explains what an airtightness test is, what the results of the tests at Moatfield Drive mean, and how architects can use airtightness tests on their projects. The results demonstrate that with some attention to detail, creative, aesthetically striking designs can meet impressive standards for airtightness. As well, they demonstrate why architects should care.
What is an airtightness test?
An airtightness test is a whole building test that measures how easy it is for air to leak through a building's enclosure or “skin.” In residential construction air leakage tests are often referred to as blower door tests, because these tests are most commonly conducted using a piece of equipment called a blower door (see Figure 1 below). A large fan or “blower” is used to extract air from or supply air to the building. For larger buildings, that is, with a floor area of over about 10, 000 ft2 (929m2), such as the OAA headquarters, this test usually requires numerous co-ordinated blower doors running at the same time. In all cases, a building must be prepared for testing beforehand by blocking intentional openings such as HVAC intake and exhaust grills, kitchen and bathroom exhaust fans, relief dampers, etc. The test equipment measures the airflow (how much air is moved into or out of the building) and the corresponding pressure difference acting across the building enclosure.
Why do you want to know about airtightness?
There are several reasons that information, especially quantitative information, about airtightness is useful and can be important.
Airtightness targets are useful during the design process for new buildings as well as retrofits – they establish quantitative expectations for a very important aspect of building enclosure performance and provide a key input into the mechanical designer’s load and energy calculations. Airtightness tests should therefore be an important part of the construction process – they provide confirmation that airtightness targets are met and, if timed properly, afford the opportunity to address problems before it is too late.
What results do you get and what do they mean?
The results of an airtightness test in raw form are the pressure difference across the enclosure, the total airflow, and the airflow direction (in or out). During testing, numerous (typically 5 to 10) airflow measurements are collected using a range of pressure differences and flow directions. This data is then plotted to quantify the relationship between airflow and pressure difference (see Figure 1 below). Experience has shown that the results tend to form a curve, and hence it is common to plot the data in logarithmic format to mathematically force the results into a straight line. The data can be presented to knowledgeable users in that format; however, it is very useful and convenient to report the flow at a single pressure from the curve and report that number.
In commercial construction, the single test pressure used for reporting is almost always 75 Pascals whereas for residential construction a pressure of 50 Pa is standard. Imperial units for air pressure, inches of water column (in H20), are sometimes used in the US.
The airflow is reported in either cubic feet per minute (cfm), liters per second (lps, or l/s), cubic meters per second (m3/s) or, in Europe, m3/hour.
It makes sense that the airflow measured will increase with the size of the building. To allow for easy comparisons between different buildings, two methods are used to normalize the airflow with respect to building size:
In single family residential construction, it has been customary to normalize by building volume. The measured airflow rate at 50 Pa is converted to air changes per hour (ACH). Air changes per hour is simply the volume of air leaked per hour, divided by the volume of the building. The test data is then reported as a number in “ACH@50”.
Air changes per hour is also sometimes used to report air leakage for commercial and larger-scale buildings. Due to the more variable surface-to-volume ratios of the different façade geometries of these buildings it is more common, and technically superior to provide the leakage rate at a specific pressure (usually 75 Pa) in terms of flow per unit area, in other words as cubic feet per minute per square foot (cfm/ft2) or litres per second per square metre (lps/m2) of building enclosure area. “Area” here means the total six sides of the cube, the four walls and sides, so all sides that are exposed to the outdoor air are part of this equation. Test results are reported in “lps/m2@75” or “email@example.com in H20”.
Like air changes per hour, this reporting method accounts for the size of the building and therefore allows you to make better comparisons between buildings. This enclosure area approach to normalizing results is increasingly finding favour in the residential community as well.
This then is how one processes the data to report it as a single number. But what does that number actually mean, in terms of a building’s performance? What is a “good” level of airtightness?
A typical new house in Ontario will have an airtightness of around 2 ACH@50 (two complete volumes of the house will leak through the enclosure every hour when a pressure difference of 50 Pascal is imposed). Older homes often reported results of 8 or 12 ACH@50. As far back as the 1980’s, the R-2000 program required airtightness to be tested and be below 1.5ACH@50. More recently, the German PassivHaus program has required airtightness of 0.6 ACH@50.
For over 20 years, the National Research Council of Canada has recommended that air leakage across the enclosure of commercial buildings be limited to a maximum of 2 lps/m2at a pressure difference of 75 Pa. In the United States, the building industry has adopted the Canadian standard and converted it to imperial units: 0.4 cfm/ft2 at 0.3”H2O. The 2 lps/m2@75 Pa and 0.4 firstname.lastname@example.org”H2O numbers are good targets for commercial building enclosures; in the United States the General Services Administration requires that all new buildings meet these targets. For higher performance buildings, the U.S. Army Corps of Engineers has a target of about 0.25 cfm/ft2 at 0.3”H2O which works out to about 1.3 lps/m2 at 75 Pa. Very high performance buildings sometimes use a target of under 1 lps/m2, but that is not always easy to reach.
Buildings tested by the Canada Mortgage and Housing Corporation, National Research Council Canada, and the US National Institute for Science and Testing have reported numbers that vary from about 1 to about 15 lps/m2 at 75 Pa. At the higher end of that scale, you would definitely know that your building is relatively leaky; as you get close to 2 or 1, you know that you have a tight building. Recent experience has shown that a deliberate plan from the start of design through to early testing during construction can routinely deliver buildings with values of under 2 lps/m2@75.
What Can Be Learned from the OAA Headquarters Tests?
The OAA headquarters are an interesting example. As mentioned above, the building was tested during a retrofit process. Many improvements had already been made (e.g. significant portions of curtain wall were replaced, many feet of caulked joints were cleaned and repaired, etc.). The goal of testing was to assess the success of these steps and to determine the need for further work in a specific area of the building.
We first tested with all intentional openings (windows, doors, mechanical penetrations) sealed. Under these conditions, airflow into the building (also called infiltration) was approximately 0.91 lps/m2 at 75 Pa. Airflow out of the building (or exfiltration) was approximately 0.92 lps/m2. So, the results show that the building was actually remarkably tight – it would more than meet the stringent target used by the U.S. Army Corps of Engineers for high-performance buildings.
The other interesting thing about the OAA results is how they show the impact of known holes. What we are finding in modern airtight buildings is that in fact the air leakage through the mechanical system, grills and openings is substantial. It's not uncommon to see half of the total leakage going through the mechanical system and that means we have to worry more about backflow dampers that actually act as backflow dampers should—that is, they should seal tight in the closed state—because this will matter to the overall tightness of the building in service.
The final test at the OAA headquarters looked at this issue of leakage through mechanical penetrations (see Figure 2 below). The test began with everything sealed, at which point the flow rate was 3.27 m3/s (6932 cfm). Note that this is a straight measure of airflow – unlike the leakage rate (lps/m2), it doesn’t account for the building’s size. So it wouldn’t work well to compare different buildings, but it does work well to compare the same building under different conditions, which is what this final set of tests did. Basically, a constant pressure of -75 Pa was maintained while the coverings of the mechanical systems were systematically removed. As soon as we opened the return air vents, there was a 41% change in flow. By the time everything was opened, the flow rate was 5.46 m3/s (11575 cfm), a 67% increase from the base rate. What this tells us is that for buildings that are already quite airtight, where it would be difficult to make further improvements through changes to the enclosure, airtightness can still be improved by fixing those mechanical leaks, for example by ensuring that supply vents in the HVAC system are actually sealed during off-cycle times.
How Do Airtightness Tests Fit into the Design Process?
The trends are clear: the building industry is going to be required to meet airtightness targets, either because they are added to building codes or energy programs or because owners are looking for better buildings – buildings with less energy consumption, more comfort, better indoor air quality, and lower risk of moisture damage. One of the ways architects can use airtightness tests is to prove to the owner and code officials that the contractors and the designer together have delivered a good, airtight building enclosure. It provides quantitative verification that the methods used were successful, much like crushing a concrete cylinder shows that the required concrete strength was achieved.
Airtightness testing can also be used diagnostically. For new construction, we typically do a test as early as possible in the construction process so that if the building fails, remedial work can be undertaken to find the cause and fix it. Similarly, if you are about to do a major energy efficiency retrofit – replace windows, add insulation or take other substantive steps – it is usually advisable to do a test before you get too far into the design of the retrofit. You want to do this as part of the assessment of the existing building and have an understanding of how leaky the building is, so that you can decide how much effort should be spent on air tightening versus how much effort should be spent on, for example, insulation or new windows.
By using airtightness tests regularly, architects and contractors can learn over time what works and what doesn’t; you can get better and better at predicting performance and therefore be comfortable designing for strong airtightness performance. Furthermore, once you actually have these targets in place, once you know that you can achieve a certain target, you can then provide input to the mechanical systems designer. As we move to tighter and tighter buildings, mechanical engineers are often left guessing about what to assume in their energy calculations for air leakage, and so they tend to aim on the safe side. If you specify an air tightness target and have a method in place to test it during construction, then the mechanical engineer can do a better job of sizing the mechanical equipment for heating, cooling, and dehumidification, and also design makeup air systems to minimize pressurization. A growing problem in airtight buildings is that mechanical engineers are over-pressurizing them and forcing air to leak out even though the building is relatively airtight. Just like manufacturers test the R- value of insulation so that you know what you're putting in the walls, and can design for it, measuring the airtightness of a building gives the mechanical engineer some values to use in system design.
Minimizing air leakage across the building enclosure is fundamental to high-performance buildings – that is, buildings that are low-energy, comfortable, healthy, and durable. In the end, airtightness testing is basically a quantitative quality control tool with the added benefit of making sure that you know what your mechanical system should be designed for.